Learning Objectives

After completing this lesson, you’ll be able to:

• Add and configure a FeatureReader.
• Use a FeatureReader with multiple initiator features.
• Add and configure a FeatureWriter.
• Use a Connector transformer.

Resources

• VancouverNeighborhoods.kml
  • C:\FMEData\Data\Boundaries\VancouverNeighborhoods.kml on Safe Software training machines
• Orthophotos.zip
  • C:\FMEData\Data\Orthophotos\*.tif
    
  • Note: you can read files from the URL if you don't have FMEData installed on your C:\ drive, but performance will be very poor. The example workspace uses local paths.
• Starting workspace
  • C:\FMEData\Workspaces\AdvancedReadingAndWriting\exercise-use-the-featurereader-and-featurewriter.fmw
• Complete workspace
  • C:\FMEData\Workspaces\AdvancedReadingAndWriting\exercise-use-the-featurereader-and-featurewriter-complete.fmw

Introduction

In this exercise, we'll look at common use cases for the FeatureReader and FeatureWriter and use a Connector transformer.

The scenario is we have a folder full of orthophotos in GeoTIFF format. However, we want to extract the orthophotos as individual rasters covering the area of several downtown neighborhoods. 

1) Start Workbench

Open the starting workspace in FME Workbench (2024.1 or later).

2) Explore the Workspace

This workspace is already mostly configured, but we need to add a FeatureReader and FeatureWriter to accomplish this task. If you examine the workspace, you'll see the following:

1. It reads a vector dataset of the neighborhoods in KML format, removes some attributes, and reprojects them.
2. We will add the FeatureReader to read the photos based on the areas of the neighborhoods.
3. It will mosaic the rasters into one image using the RasterMosaicker.
4. Then it will clip the neighborhood features out to create individual raster features using the Clipper.
5. Then it will process the rasters to set a NoData band so the areas outside the photo are transparent.
6. Then we need to configure a FeatureWriter to write out the individual raster files.
7. Then we can use a Connector transformer to upload the files to a web storage location.

3) Add a FeatureReader

Add a FeatureReader to the Orthophotos bookmark above the Dissolver. Connect the Dissolver:Area port to the FeatureReader Initiator port.

Double-click the FeatureReader to open its Parameters dialog. Configure it as follows:

With this configuration, the FeatureReader will read the GeoTIFF files that overlap the neighborhoods of interest.

In this case, we want to read each photo that intersects any neighborhood and just read it once. That is why we used a Dissolver to get a single initiator feature. But the FeatureReader can also be used in scenarios with multiple initiators. If we skipped the Dissolver in this case, we'd read some of the photos multiple times, as we would read intersecting photos for each neighborhood feature separately. Try that out - you'll need the number of features to answer a Quiz question below.

Click OK.

Use Run to This on the FeatureReader. Inspect the GEOTIFF output port to view the matching orthophotos. Take note of how many there are; you'll need it for the quiz. The results should look like this:

Map tiles © Stadia Maps, © OpenMapTiles, © OpenStreetMap contributors, © Stamen Design

4) Run the Workspace

Seeing that the FeatureReader is configured properly, connect the FeatureReader's  GEOTIFF output port to the RasterMosaicker input port.

Run the workspace to update the feature caches. Inspect the RasterBandNoDataSetter Output port. You should see the images from earlier but now clipped to the neighborhood outlines and with transparent pixels for the NoData values.

Map tiles © Stadia Maps, © OpenMapTiles, © OpenStreetMap contributors, © Stamen Design

Note that raster images can take a few seconds to show up in Visual Preview.

5) Add a FeatureWriter

Add a FeatureWriter and connect the RasterBandNoDataSetter Output port to the Connect Input port:

It will turn into a port named Output, the default name of the feature type it will create.

Double-click the FeatureWriter to open its Parameters. Configure them as follows:

With this configuration, the FeatureWriter will use a feature-type fanout to create a separate GeoTIFF file for each neighborhood.

Click OK.

You'll see the name of the port change again, this time to @Value(NeighbohoodName), reflecting the feature type fanout setting.

Run your workspace to write out the images. Click the FeatureWriter once to select it and click the Open Containing Folder button to see the written images:

6) Add an FTPCaller Transformer

Alternatively, you can try using a different Connector transformer with a web service you have access to, like a GoogleDriveConnector or DropboxConnector.

The FeatureWriter is working properly to write the data, but now it's time to take advantage of being able to continue a workspace after writing. We can use a Connector transformer, or in this example, an FTPCaller, to upload these files somewhere.

Click the Summary cache to inspect it. You should see an attribute called _dataset. This attribute was created by the FeatureWriter and stores the path to the written data. You can use this path to find the files that were just written and do something with them; in this case, upload them to an FTP server.

Add an FTPCaller after the FeatureWriter and connect it to the Summary port.

Once you have a Web Connection configured, double-click the FTPCaller to open its parameters. Configure it as follows:

Click OK.

Run your workspace. In the log file, you will find the confirmation that the dataset was uploaded to the FTP Server:

Here is the key advantage of the FeatureWriter: your workspace can continue after you write data.